Security Sales & Integration May 2023 - 27

Improving Pattern Recognition
Every year, improvements are made to AI algorithms.
These computer programs are presented
with, in this case, images containing
examples of target concepts such as " blurry
image " or " blue car. " The AI picks out statistical
patterns in the images representing
those target concepts.
Each year, AI's ability improves in its ability
to cue in on true patterns and filter out spurious
or random variations that sometimes are part of
the target concept, such as reflections, shadows
or obstructing objects. We use AI based on patterns
and predictions because there is too much
data to sift through, and it is too complex for
people to do so manually.
Another important contributing factor is the
computing hardware used for artificial intelligence.
Algorithms are instructions, much like a
recipe, before a chef cooks a meal. AI is an algorithm
that writes its own rules from example
images provided by the user in a process called
" training. " The AI can then use those rules on
new data it has not seen before, to automatically
do the given task.
The hardware used to execute the rules of an
AI algorithm has advanced significantly, too, in
terms of more computations per second. This
new hardware enables AI algorithms to find
better and more authentic patterns in ever larger
datasets of examples. Developers train AI
software to recognize specific objects or individuals,
allowing it to filter out irrelevant
information and focus on important events.
The more accurate an AI's functionality
is, the more it becomes possible to monitor
increasingly larger areas with fewer
cameras and personnel, reducing costs and
increasing coverage.
Maximizing Safety & Efficiency
It's crucial to find applications contributing to
a location's safety. Shoppers in a mall, for example,
expect protection from security cameras
placed throughout the building. People may let
their guard down in these secured areas because
of the perceived safety.
If the shopper is aware that AI software is installed
to augment the capabilities of video surveillance
devices, security managers and shoppers
will have the assurance that if an incident
occurs, such as a thief stealing someone's purse,
the software will confirm that the camera was
recording properly at the time of the incident.
securitysales.com
MAY 2023 Security Sales & Integration 27
AI software will also realign security guards'
and property managers' expectations. If a security
guard has AI software monitoring their surveillance
devices, it can simplify their job duties
in several ways.
AI software can automatically analyze and
process large amounts of data from the surveillance
cameras in real-time, which means the
guard doesn't have to monitor the live camera
feeds manually and continuously. The security
managers are enabled through the software to
know that each unit is working correctly. This
advancement frees the guard's time to focus on
other essential tasks, such as conducting patrols
or responding to alarms.
Better outcomes are the result of better performance,
and this sentiment remains true for
systems integrators looking to add AI software
to their security cameras. Security managers
will achieve optimal video surveillance with
AI software because this technology maximizes
safety and efficiency. As a byproduct, the cost
will decrease, making effective and efficient video
surveillance equipped with AI software more
prevalent across industries.
Every year improvements are
made to AI algorithms.
In 2023, the video
surveillance
industry is
anticipated to
be worth roughly
$62.6 billion.
The more accurate an AI's
functionality is, the more it
becomes possible to monitor
increasingly larger areas with
fewer cameras and personnel,
reducing costs and increasing
coverage.
MOHAMAD ZAKI/STOCK.ADOBE.COM
GORODENKOFF/STOCK.ADOBE.COM
http://www.securitysales.com
Security Sales & Integration May 2023

Table of Contents for the Digital Edition of Security Sales & Integration May 2023
